@Ylecun perpetuates false dichotomy here, between nature & nurture, without presenting data. Is there any evidence primates have fewer innate priors than mountain goats? Any proof that more innate priors actually entails less adaptivity? Better priors can yield *better* learning.https://twitter.com/ylecun/status/1180171721481113602 …
-
I believe
@ylecun is on the record advocating for the *minimal* set of maximally powerful built in structure (e.g., convolution being one)
in his debate with me he advocated for zero, not even convolution. the question is how minimal the minimal set should be, ala einstein’s simple but not simpler.